15-462: Computer Graphics Nancy Pollard Assistant Professor Robotics Institute and Computer Science Department Introduction. Administrivia. Who am I?. What will we do in this course?. What is Computer Graphics?
Krishnamurthy,Sandeep, William P. Bottom and Ambar G.
Rao(2003), 'Adaptive Aspirationsand Contributions to a Public Good: Generic Advertising as a Response toDecline', Organizational Behavior and Human Decision Processes (OBHDP),92(1/2), 22-33.AbstractGeneric advertising,the promotion ofan entire product or service category,is a common form of inter-firmcooperation. Krishnamurthy,Sandeep(2001), “The Impact of Provision Points on Funding Generic AdvertisingCampaigns”, Marketing Letters, 12(4), 315-325.AbstractGeneric advertising promotes thegeneral qualities of a product category and, therefore, benefits all firms inthe category. Recent examples of such campaigns organized through voluntarycontributions by firms are found in the Life Insurance, ('It's 1998.You're Dead. What Do You Do Now?' ) and Steel industries ('The NewSteel. Feel the Strength.' Free-riding is commonly observed insuch campaigns when they use the Simple Voluntary Contribution Mechanism (VCM)-here the budget is obtained by simply adding up the voluntary contributions offirms.
We propose a new mechanism called the Provision Point VCM that can helpalleviate this free-riding problem. Here, a target budget called the provisionpoint is announced with the condition that the campaign will be mounted if andonly if the contributions equal or exceed this point. First, analytically webriefly describe the results for the Simple VCM. Here, free-riding will alwaysresult in equilibrium and the advertising budget never maximizes industrywelfare. Then, we show that if the provision point is set equal to the ParetoOptimum, the industry welfare maximizing advertising budget is always afeasible equilibrium outcome and is a unique outcome in some cases. In fact, weshow that no other Provision Point VCM can lead to welfare maximization. Then,we present results from experiments using MBA students that test thesefindings.
Consistent with our analytical results, we show that when theprovision point is set equal to the Pareto Optimum, it outperforms the SimpleVCM in terms of overall contributions and free-riding behavior. Krishnamurthy,Sandeep (2001), 'Communication Effects In Public Good Games With AndWithout Provision Points', Research In Experimental Economics,Volume Eight, Editor: Mark Isaac, JAI(An Imprint of Elsevier Science B.V, Amsterdam,The Netherlands).AbstractIn this paper, I report the resultsof thirty economic experiments investigating if the short-term and long-termimpact of face-to-face communication is different in provision point and simplevoluntary contribution mechanisms. This is an interesting issue in theexperimental economics community because technically such communication isnon-binding and should not impact the equilibrium prediction (i.e., observedbehavior). My results show that communication does have short-term andlong-term (it impacts contributions in periods immediately following it inwhich no communication is allowed) effects. However, in provisionpoint mechanisms, the long term impact of communication is strikingly strongerand the short-term impact is affected more by the history of past interaction.Communication leads to a drop in the number of free-riders and cheap-riders inboth mechanisms. I statistically test twenty seven hypotheses and findstrong support for twenty three of them. I also investigate if the nature ofcommunication has an impact on contributions.
I find that, ceterisparibus, the magnitude of communication (e.g. The number of words spoken) isnegatively correlated with efficiency. I found no significant correlationbetween the distribution pattern of communication within a group andcontributions. Krishnamurthy, Sandeep (1999),“Enlarging the Pie vs. Increasing One’s Slice: An Analysis of the RelationshipBetween Product Class and Brand Advertising”, Marketing Letters, 11(1),37-48. Abstract Generic advertising promotes the generalqualities of a product category and, therefore, benefits all firms-regardlessof who has contributed to the campaign (e.g. Suchcampaigns are organized either by independent contributions by industry membersor through government legislation.
In this paper, we study the relationshipbetween brand and generic advertising in these two cases. In theindependent-contribution case, a free-riding or cheap-riding genericadvertising equilibrium is predicted. Interestingly, in the free-ridingequilibrium, we show that dominant firms in industries should be indifferent tofree riding by lesser firms. They should incur the entire industry advertisingexpense and be better for it. In the cheap-riding equilibrium, a group ofequally dominant firms foot the bill. In the government-sponsored case, weestablish that industry spending on generic advertising is greater.
But, wefind that there is an increase in total spending on brand advertising as well. Knowledge Areas OPEN SOURCE Papers. Krishnamurthy,Sandeep (2005), “On the Intrinsic and Extrinsic Motivation of Open SourceDevelopers”, Forthcoming in Knowledge,Technology & Policy.AbstractMotivationin the context of open source software may be seen as fundamentally differentdue to the presence of unpaid programmers, implicit rather than explicit formsof control and a different methodology for software development. Since software development is a creativetask, the motivation of open source programmers can be compared to individualsin creative industries (Caves 2002). This paper summarizes the important trends in the research on motivationin open source and identifies variables that should be included in futureresearch. Specifically, the currentliterature favors a taxonomy that considers two components of motivation-intrinsic (e.g.
Fun, flow, learning, community) and extrinsic (e.g. Financialrewards, improving future job prospects, signaling quality).
I make a case for incorporating bothelements in developing an integrative theory about developer motivation. Three elements are identified as beingunique to FLOSS development- diversity of project structures, co-existence ofcompanies and communities and co-existence of creative and commercialelements.
The important empiricalevidence on FLOSS developer motivation is presented and analyzed. Four factors are identified as importantmitigating and moderating factors in the conversation surrounding developermotivation- financial incentives, nature of task, group size and groupstructure. The role of these factors ondeveloper motivation is discussed. Krishnamurthy,Sandeep (2005), “About Closed-door Free/Libre/Open Source (FLOSS) Projects:Lessons from the Mozilla Firefox Developer Recruitment Approach”, Upgrade, Volume 6, Issue 3, Availableat.AbstractIn thispaper, the notion of a 'closed-door open source project' is introduced.In such projects, the most important development tasks (e.g. Code check-in) arecontrolled by a tight group. I present five new arguments for why groups maywish to organize this way. The first argument is that developers simply do nothave the disposable time to evaluate potential members.
The next two argumentsare based on self-selection- by setting tough entry requirements the projectcan ensure that it gets high quality and highly persistent programmers. Thefourth argument is that expanding a group destroys the fun. The fifth argumentis that projects requiring diverse inputs require a closed door approach. Krishnamurthy,Sandeep and Arvind Tripathi (2005), “Bounty Programs in Free/Libre/Open SourceSoftware (FLOSS): An Economic Analysis”, Forthcoming in The Economics of Open Source Software Development, Editors- JurgenBitzer and Phillip Schroeder, Elsevier Publications.AbstractThemotivation of FLOSS developers has been an enduring topic of research. The current scholarly thinking is that themotivation of a developer could be intrinsic (e.g. Fun, creativity, flow) orextrinsic (i.e., tangible rewards such as maximization of future careerprospects).
In this paper, we focus onone mechanism for providing extrinsic financial rewards- bounty programs. Bounty programs are commonly used in FLOSScommunities to motivate developers to create new programs, to improve thesecurity of an existing product, to solve a specific bug, to obtain productimprovement suggestions, to help maintain the codebase and to preparedocumentation. We start by defining sixcharacteristics of FLOSS bounty programs- financial benefit, task-specific,competitive, judgment by expert panel, deadline, public and open source. Bounty programs can be formal or informal, maybe funded by individuals, corporations or non-profit organizations and mayprovide small to very large amounts. Thewinner-take-all feature of most bounties places undue pressure on developersand judges. The rational response to abounty can be best modeled using Selby Jr.
And Beranek(1981)’s seminalpaper. Entry is determined by acost-benefit analysis. Six potentially negative impacts on the softwaredevelopment process are identified.
Future research is needed to understand the impact of bounty programs onFLOSS in general. Krishnamurthy,Sandeep (2003), 'A Managerial Overview of Open SourceSoftware', Business Horizons, 46(5),September-October, 47-56.AbstractOpen-source software(OSS) programs(e.g. LINUX, Apache) provide access to the source code to any interestedparty. This leads to a distributed innovation model where users activelyparticipate in the development of the product.
The main purpose of thispaper is to help managers who have to choose between OSS and commercialsoftware. Not all OSS products are free. OSS products aredistributed under a wide variety of public licenses.
They are morereliable, provide greater flexibility and choice to the user and are frequentlyfree. On the other hand, OSS leads to a proliferation of versions and mayappeal only to the high-end user.
Vendors of open-source products makemoney by selling the program on a CD, providing support services to enterprisesand authenticating versions. The open source system leads to fascinatingcompetitive and cooperative relationships among companies, between a companyand a community and among communities.
Krishnamurthy,Sandeep(2005), 'An Analysis of Open Source Business Models',Forthcoming in Making Sense of the Bazaar: Perspectives on Open Source andFree Software, Editors- Joseph Feller, Brian Fitzgerald, Scott Hissam and Karim Lakhani, MIT Press, Boston, MA.AbstractOpen-source software is not forhobbyists any more. Instead, it is a business strategy with broadapplicability.
Businesses can be built around this idea. In thispaper, I want the reader to grapple with the specifics of how to build and growsuch a business.
To this end, I have proposed three fundamental businessmodels- Distributor, Software producer GPL and non-GPL and the Third-PartyService Provider. These are sustainable models that can lead to robustrevenue streams. The business models provided here can be enhanced by theaddition of further revenue streams. For instance, we now know thatcertification of developers on an Open-Source product can lead to strongrevenues. Not all products have the same profit potential. Therefore,not all Open Source Software products have the same profit potential.
Ihave classified Open Source Software products into four categories- Stars,High-profile nichers, Low-profile nichers and Mainstream utilities.Businesses can be built around Stars. High-profile nichers can lead torobust revenue streams if properly marketed. The other two categories maynot lead to high profits. Since many Open Source Software products arefreely available, managers must scan public repositories to find out whichproducts will be suitable for their business. The future of Open SourceSoftware is bright. Increasingly, we will find that these products willtake a central role in the realm of software and will find a larger place inall our lives. Krishnamurthy,Sandeep(2002), 'Cave or Community?
An Empirical Examination of 100 MatureOpen Source Projects', First Monday, 7(6).AbstractStarting with Eric Raymond'sgroundbreaking work, 'The Cathedral and the Bazaar', open-sourcesoftware ( OSS ) has commonly been regarded as work produced by a communityof developers. Yet, given the nature of software programs, one also hears ofdevelopers with no lives that work very hard to achieve great product results.In this paper, I sought empirical evidence that would help us understand whichis more common - the cave (i.e., lone producer) or the community. Based on astudy of the top 100 mature products on Sourceforge, I find a few surprisingthings. First, most OSS programs are developed by individuals, rather thancommunities. The median number of developers in the 100 projects I looked atwas 4 and the mode was 1 - numbers much lower than previous numbers reportedfor highly successful projects! Second, most OSS programs do not generate a lot ofdiscussion. Third, products with more developers tend to be viewed anddownloaded more often.
Fourth, the number of developers associated with aproject was positively correlated to the age of the project. Fifth, the largerthe project, the smaller the percent of project administrators.Knowledge Areas Spam/Permission Marketing PapersPapers. Krishnamurthy,Sandeep(2002), 'The Ethics of Conducting E-Mail Surveys', Forthcomingin Readings in Virtual Research Ethics: Issuesand Controversies, Edited by Elizabeth A. Hershey, PA: Idea Group.AbstractThe prospect of using e-mail insurvey research can be very exciting to academic researchers. However, itraises many ethical concerns.
While many people have started to say thatobtaining consumer permission is important, there is no clarity on how toobtain and maintain permission. Some academic researchers might arguethat due to the low volume and infrequent nature of their surveys and thegeneral positive perception of academia, their e-mail surveys do not add to theSpam problem. However, this is problematic from an ethical perspectivesince it changes the definition of what Spam is from any unsolicited e-mail toa subset of these e-mails which have certain predefined characteristics.There are ways to implement permission-based respondent contact if the academiccommunity wants to. The only negative to keep in mind will be thestatistical problem of self-selection and the 'loss of completerandomness' to some degree.
Regardless, the future legal landscapemay force academic researcher to adopt permission as the standard. Krishnamurthy,Sandeep (2001), “A Comprehensive Analysis of Permission Marketing”, Journalof Computer-Mediated Communication, 6(2 ).Abstract Godin (1999) has proposed a new idea-permission marketing. Here, consumers provide marketers with the permission tosend them certain types of promotional messages. This is seen as reducingclutter and search costs for the consumer while improving targeting precisionfor marketers. This paper makes three contributions: First, a critical analysisof the concept and its relationship to existing ideas in the marketingliterature is discussed. Second, a taxonomy of four models used to implementpermission marketing today, direct relationship maintenance, permissionpartnership, ad market and permission pool, is presented.
Permission intensityis seen as a key differentiator among models. Finally, a comprehensiveconceptual cost-benefit framework is presented that captures the consumerexperience in permission marketing programs. Consumer interest is seen as thekey dependent variable that influences the degree of participation.
Consumerinterest is positively affected by message relevance and monetary benefit andnegatively affected by information entry/modification costs, message processingcosts and privacy costs. Based on this framework, several empirically testablepropositions are identified. Krishnamurthy, Sandeep (2000), “ Spam Revisited ”, Quarterly Journal of ElectronicCommerce, 1(4), 305-321. AbstractEven though unsolicited commerciale-mail or Spam continues to be a major problem, very little academic researchhas focused on it. Notable exceptions include Shiman(1996), Cranorand LaMacchia(1998), Samoriski(1999) and Sheehan and Hoy(1999). The purpose ofthis paper is to provide a comprehensive overview of the Spamproblem and a critical analysis of the solutions. We begin with a definition ofSpam.
This is followed by an analysis of the pernicious impact of Spam on themajor stakeholder groups- consumers, Internet Service Providers,legitimate advertisers, e-commerce firms and employers. Since there is no“clean” solution to limiting the volume or nature of Spam, avariety of responses to Spam exist. These can be placed into four categories-laissez faire arguments, business-initiated solutions, third-partyoversight, consumer education and legislative solutions. For the remainder ofthe paper, we focus upon one business-initiated solution- permissionmarketing (Godin 1999). Permission marketing envisages a world where consumerscontrol the promotional messages targeted at them.
Consumers provide a firminformation about their interests and product preferences. The firm then sendsthe consumer promotional messages based on this information.
We argue that, eventhough this idea has merit, there has been an inconsistent application of thisidea. To support this, we discuss six variants of the original concept of Spamthat purport to use permission marketing. Based on thisdiscussion, we end by identifying six key elements that must be a part of everypermission marketing campaign if it has to be clearlydistinguished from Spam. Krishnamurthy,Sandeep (2000), Review of Seth Godin's 'Permission Marketing: TurningStrangers into Friends and Friends into Customers', 37(4), Journal ofMarketing Research, 525-528. Book reviewNo Abstract ProvidedPrivacy Papers Miyazaki, Anthony and Krishnamurthy,Sandeep(2002), “ Internet Seals of Approval: Effectson Online Privacy Policies and Consumer Perceptions ”, Journal of Consumer Affairs, 36(1), Summer, 28-49.The use of Internetseal of approval programs has been touted recently as an alternative topotential legislation concerning consumer-related online privacy practices.Questions have been raised, however, regarding the effectiveness of suchprograms with respect to maintaining privacy standards and aiding onlineconsumers. The authors examine these issues in a series of four studies, thefirst of which examines Internet seal of approval logo usage in the currentmarketplace.
The next study applies Federal Trade Commission privacy standardsto various online privacy policies in an effort to determine the ability ofseal of approval program participation to act as a valid cue to a firm™s stateprivacy practices. The last two studies are experiments designed to ascertainhow online firm participation in Internet seal of approval programs affectconsumers. Implications for policy and industry are also discussed. Krishnamurthy, Sandeep and Nitish Singh, “The International E-MarketingFramework (IEMF)- A Guiding Template for Future Global E-MarketingResearch”,Forthcoming in International MarketingReview.Purpose:International E-Marketing is emerging as an important area for marketers,as global online markets expand. This special issue is an attempt to encourage,showcase,and guide research in the area of International E-Marketing.Research/PracticalImplication: In the editorial, we introduce the InternationalEmarketing Framework (IEMF) as a guiding template for future research ininternational e-marketing.
The IEMF should help shape scholarly inquiry in thedomain of international E-Marketing, classify current intellectualcontributions in this area and delineate the gaps in the literature.Originality/Value:The editorial presents the International E-Marketing Framework andclassifies various papers in this issue using this framework. Finally, theeditorial concludes with several compelling research questions to motivatefuture research in this area. Krishnamurthy, Sandeep (2005), “Introducing E-MARKPLAN- APractical Methodology to Plan E-Marketing Activities”, Business Horizons, Forthcoming.Even though E-Marketing is highly prevalent, there is no template formanagers who wish to use the Internet/Web and related information technologiesto market their products and services.The purpose of this paper is to provide managers with a comprehensive,actionable, and practical methodology (E-MARKPLAN) to plan, enact, and analyzeE-Marketing activities. Five case studies are used to illustrate the diversityof E-Marketing actions. E-MARKPLAN consists of five parts- Goals, Actors (i.e.,those who take E-Marketing actions), Spaces (i.e., theaters of engagement),Actions, and Outcomes.
E-MARKPLAN is versatile and is not limited to companieswho have an Ecommerce operation. Krishnamurthy,Sandeep(2003), 'A Comparative Analysis of Amazon and eBay',Forthcoming in Intelligent Enterprises Of The 21st Century, Editors:Jatinder N. Gupta and Sushil K. Sharma, Idea Group Publishing,Hershey, Pennsylvania, USA.AbstractEven though Amazon.com has receivedmost of the hype and publicity surrounding E-Commerce, eBay has quietly builtan innovative business truly suited to the Internet. Initially, Amazonsought to merely replicate a catalog business model online. Itstechnology may have been innovative- but its business model was not. Onthe other hand, eBay recognized the unique nature of the Internet and enabledboth buying and selling online with spectacular results.
Its auctionformat was a winner. EBay also clearly demonstrated that profits do not have tocome in the way of growth. Amazon was intially focused on BN.com as acompetitor.
Over time, Amazon came to recognize eBay as the competitor.Its initial foray into auctions was a specacular failure. Now, Amazon istrying to compete with eBay by facilitating selling and strengthening itsaffiliates program. Krishnamurthy,Sandeep (2001), “Understanding Online Message Dissemination: An Empirical Analysisof Send-This-Message-To-Your-Friend Data”, First Monday, 6(5).AbstractOur goal in this paper was toanalyze 'send-this-story-to-a-friend' data.
Encouraging consumers tosend stories to friends is an application of viral marketing which envisionsmessage dissemination through customer-to-customer communication rather thanfirm-to-customer communication. It makes sense for a publisher to include thisfeature since individuals are more likely to accept the recommendations oftheir friends and this may increase theit probability of reading the story bymaking it more salient. We collected publicly available data from ESPN.com overan 11-day period. Using this data, our findings are: first, the number of timesstories are sent out represent a miniscule proportion of visits to the site;second, if the top story is very influential, it may increase the total visitsto the site and there may be spillover effect; third, there is an interestingweekend effect with a drop in the number of stories sent out; and, finally weobserve an interesting floor effect when we analyze the ratio of the top andbottom stories.
Krishnamurthy,Sandeep (2001), “Person-to-Person Marketing: Marketing and the New ConsumerWeb”, Quarterly Journal of Electronic Commerce, 2(2), 123-138.AbstractWe propose a new form of marketingcalled, person-to-person(p2p) marketing. This is characterized bythree elements- consumer collectives (both private social networks and publiccommunities) conducting tasks usually performed by firms, diminished firmcontrol on the marketing process and customer empowerment due to the end of isolation.We see the impact of consumer collectives in virtually all marketing processes-product development(e.g. Linux, Slashdot), message dissemination, i.e., viralmarketing(e.g. Hotmail, Paypal), product evaluation(e.g. EPinions), (digital)product sharing(e.g. Napster, Gnutella), product purchase(e.g.
EBay, Mobshop)and customer feedback(e.g. Our main objective in this paperis to identify P2P marketing as a major event in the marketing landscape and toplace its impact in the context of the literature. Krishnamurthy,Sandeep (2001), “An Empirical Study of the Causal Antecedents of CustomerConfidence in E-tailers”, First Monday, 6(1).Abstract Why is it that consumers are very confident with an e-tailer such as Amazon.com and lack the same confidence when it comes to a smaller e-tailer such as supremevideo.com?
In this paper, we attempt to answer this important question by examining the antecedents to customer confidence in e-tailers, using secondary data. Our findings indicate that the ease of use of a site, the level of online shopping resources, and the presence of a trusted third party seal all positively impact the level of customer confidence. Interestingly, online relationship services did not have an impact on consumer confidence. Larger firms may have a small edge. We also find that there are no large differences in the results across different product categories. Krishnamurthy,Sandeep (2000), “ Deciphering the Internet AdvertisingPuzzle ”, Marketing Management, 9(3),34-39.AbstractInternet advertising presents apuzzle to advertisers. On the one hand, total spending on Internet advertisinghas exploded and everybody (including your competition) seems to be advertisingon it.
On the other, click-through rates on banners have dropped to about 0.5%and the most basic measures of ad viewership and effectiveness have been calledinto question. In this article, our goal is to provide you with some guidelineson how to think of Internet advertising. We summarize the measurement problemswith banners and introduce you to the concepts of permission and viralmarketing. We end with six prescriptions to keep in mind when you are puttingtogether your campaign.
The Internet is evolving rapidly and managersinterested in advertising on it must be prepared to adapt in real-time if theyare to be successful. Krishnamurthy,Sandeep (2001), “ The Microsoft Challenge Case ”, International Journal of Non-profit and VoluntarySector Marketing, 6(2), 105-115.AbstractGraduate students are integral tothe mission of a research university.
Yet, due to increased numbers of suchstudents and the decreased number of jobs available to them, they face hugefinancial challenges. Universities have been unable to fund all theirgraduate students. In this case, I propose that striking creativealliances with for-profit corporations is one avenue to meet these fundingneeds. Specifically, I discuss the Microsoft Challenge- a unique arrangementwhere Microsoft challenged the faculty at the University of Washington to pay for graduate students.I present the rules and the results of this challenge. Lessons learnedfrom this challenge have important implications for all universities faced withpressures to fund graduate students.
Krishnamurthy,Sandeep(2002), “ The Washington Combined Fund Drive Case ”, International Journal of Non-profit and VoluntarySector Marketing, 7(1), 45-55.AbstractThis is the case of a largeworkplace giving program- the Washington Combined Fund Drive. This program is different inmany ways from most workplace giving programs- it provides individuals a largechoice of charities(e.g. Single charities, federations, United Ways) andpayment options(e.g. Check, payroll deduction). In most workplace givingprograms, United Way has strong monopoly power. Extensive data on theperformance of the program is provided. The program has grown steadilyand is now leveling off.
The director of the program is trying toidentify the most effective way to improve the program.
Since a computer can ultimately deal only with binary numbers, pictures have to be DIGITIZED (reduced to lists of numbers) in order to be stored, processed or displayed. The most common graphical display devices are the CATHODE RAY TUBE inside a monitor, and the PRINTER, both of which present pictures as two-dimensional arrays of dots. The most natural representation for a picture inside the computer is therefore just a list of the colours for each dot in the output; this is called a BITMAPPED representation. Bitmapping implies that each dot in the picture corresponds to one or more bits in the computer's video memory.An alternative representation is to treat a picture as if it were composed of simple geometric shapes and record the relative positions of these shapes; this is called a vector representation (see more under VECTOR GRAPHICS).
A typical vector representation might build up a picture from straight lines, storing only the coordinates of the endpoints of each line.Vector and bitmapped representations have complementary strengths and weaknesses. Most modern s work by drawing dots, which makes it more efficient to display bitmapped images. There are a few, mostly obsolete, display technologies, such as the GRAPH PLOTTER or the VECTOR DISPLAY tube, that can draw lines directly, but it is more typical for a vector image to be first converted into bitmap (the process called RENDERING) before displaying it on a dot-oriented device. Rendering involves extra work for the computer, which is performed either by software or by special hardware called a GRAPHICS ACCELERATOR.Vector representations are easily edited by moving, resizing or deleting individual shapes, whereas in a bitmap all that can be changed is the colour of the individual pixels.
Vector images can easily be rotated and scaled to different sizes with no loss of quality, the computer simply multiplying the endpoint coordinates by a suitable factor. A bitmap, on, the other hand, can be magnified only by duplicating each pixel, which gives an unsightly jagged effect.Vector representations are most suitable for pictures that are actually drawn on the computer (such as engineering drawings, document layouts or line illustrations), and for images that must be reproduced at various sizes (such as fonts). On the other hand, bitmaps are more suitable for manipulating photographs of real world scenes and objects that contain many continuously varying colours and ill-defined shapes - both scanners and digital cameras produce bit-mapped images as their output.